The consideration of those who died before the age of marriage properly comes under the head of lethal selection, but if attention is confined to those who, though reaching the age of marriage, fail to marry, sexual selection still has importance. For instance, it is generally known (and some statistical proof will be given in Chapter XI) that beauty is directly associated with the chance of marriage. The pretty girls in general marry earlier as well in larger percentage; many of the ugly ones will never find mates. Herbert Spencer argued ingeniously that beauty is associated with general mental and moral superiority, and the more exact studies of recent years have tended to confirm his generalization. A recent, but not conclusive, investigation[65] showed beauty to be correlated with intelligence to the extent of .34. If this is confirmed, it offers a good illustration of the action of sexual selection in furthering the progressive evolution of the race. Miss Gilmore, studying a group of normal school graduates, found a direct correlation between intelligence (as judged by class marks) and early marriage after graduation. Anyone who would take the trouble could easily investigate numerous cases of this sort, which would show the effect of sexual selection in perpetuating desirable qualities. But sexual selection no longer has the importance that it once had, for nowadays the mere fact of marriage is not a measure of fecundity, to the extent that it once was. In the old days of unlimited fecundity, the early marriage of a beautiful, or intelligent, woman meant a probable perpetuation of her endowments; but at present, when artificial restraint of fertility is so widespread, the result does not follow as a matter of course: and it is evident that the race is little or not at all helped by the early marriage of an attractive woman, if she has too few or no children. Fecundal selection, then, is becoming the important phase of reproductive selection, in the evolution of civilized races. The differential birth-rate is, as we have often insisted, the all-important factor of eugenics, and it merits careful consideration from all sides. Such consideration is made difficult by the inadequate vital statistics of the United States (which ranks with Turkey and China in this respect); but there is no doubt that the birth-rate as a whole is low, as compared with that of other countries; although as a whole it is not dangerously low and there is, of course, no necessary evil in a low birth-rate, of itself, if the quality be satisfactory. The U. S. Census tabulation for 1915 gives the following comparison of the number of babies born alive each year, per 1,000 population, in various countries: